Accurate prediction of treatment response and selection of optimal treatments remain challenging in epilepsy management. With no reliable surrogate biomarkers for treatment response, the current process of selecting an antiseizure medication remains largely a trial-and-error approach. Other non-pharmacological treatment options, such as epilepsy surgery, are viable alternatives for patients with drug-resistant epilepsy. Statistical and machine learning techniques have been used to predict seizure outcomes associated with antiseizure medications and epilepsy surgery. Recent breakthroughs in deep learning have unveiled new pathways and opportunities, potentially revolutionising personalised treatment selection in health care. In this Review, we explore a broad range of studies that have used various statistical and machine learning methodologies, with particular emphasis on state-of-the-art deep learning techniques to predict the outcomes of both pharmaceutical and surgical treatments for epilepsy. We also review potential future research trajectories and address the inherent challenges of incorporating machine learning into the clinical management of epilepsy.
Purpose Tendon transfers and nerve transfers are two techniques available for reconstruction of elbow extension, grip, pinch, and release after spinal cord injury. This systematic review was aimed at assessing and comparing the strength outcomes of tendon and nerve transfers for functional reconstruction of the upper extremity to better guide surgical discussion. Methods This review was registered with PROSPERO ID: CRD42024595503 and followed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines. Embase, Scopus, and MEDLINE were queried for citations that presented strength outcomes for patients having undergone tendon and/or nerve transfers to restore upper-extremity function after a spinal cord injury. Weighted averages were used to report patient demographic data. A pooled estimate of strength outcomes was completed for the brachialis-to-anterior interosseous nerve transfer using a random-effects model. Results Five hundred and ninety-four citations were screened for 51 included studies with a total of 959 patients who were, on average, 31.8 years old (range, 5−76). For elbow extension reconstruction, strength of M3 or greater was reported in 68% of patients who underwent nerve transfers, 79% of patients who underwent a biceps-to-triceps tendon transfer, and 70% of patients who underwent a deltoid-to-triceps transfer. The pooled proportion of patients achieving M3 or better strength after a brachialis-to-anterior interosseous nerve transfer was 0.5 (0.31−0.7), with better outcomes in younger patients. For the supinator-to-posterior interosseous nerve transfer, 78% of patients reported achieving M3 strength or better. Conclusions There is no gold standard for the functional reconstruction of the upper extremity after a spinal cord injury. The choice of procedure(s) remains patient-specific. Several options are available to surgeons treating tetraplegia, and patients have achieved considerable strength recovery from both nerve and tendon transfers. Current evidence demonstrates poor outcomes after brachialis-to-anterior interosseous nerve transfers. More homogeneous follow-up protocols are needed for strength measurement and patient-reported outcome measures for both nerve and tendon transfers. Type of study/level of evidence Therapeutic III.
Objective Cognitive impairment is cited as one of the most consequential comorbidities of epilepsy. Cognitive difficulties may be managed using compensatory strategies. However, little is understood about how these strategies are used by adults with epilepsy. The Everyday Compensation (EComp) Questionnaire offers a means of assessing the use of these strategies but has only been validated for neurologically healthy older adults. The present study examined the preliminary psychometric properties of the EComp Questionnaire in epilepsy. Methods 37 adults with focal epilepsy participating in a randomized controlled trial of a cognitive prehabilitation program completed the EComp Questionnaire, Quality of Life Epilepsy Inventory-31 (QOLIE-31), and the Instrumental Activities of Daily-Living Scale (IADL-C) at baseline and follow-up. Analyses included evaluation of internal consistency, construct validity, test–retest reliability, and item-level performance. Results Across 20 imputed datasets, the EComp Questionnaire demonstrated acceptable Cronbach’s α values. Test–retest ICCs were moderate (0.50–.58) but imprecise, with 95 % CIs spanning poor to good reliability. Correlations between the EComp Questionnaire and the IADL-C (ρ = 0.34, p = 0.04) and QOLIE-31 (r = -0.43, p = 0.007) suggested support for construct validity. Select items demonstrated low endorsement and weak association with the total score. Significance Findings provide preliminary support for the EComp as a tool for assessing compensatory strategy use in adults with epilepsy. Understanding these adaptive behaviors and functional needs can assist providers in formulating treatment plans and interventions. Future research with larger and diverse epilepsy samples will be necessary to confirm structural validity.
Background Use of antibiotic prophylaxis (AP) prior to invasive dental procedures (IDPs) to prevent infective endocarditis (IE) fell dramatically in the UK when the National Institute for Health and Care Excellence (NICE) recommended its cessation in 2008. However, guidance implementation by dentists, general medical practitioners (GPs) and others differed in England and Scotland, particularly after 2018 when the Scottish Dental Clinical Effectiveness Programme advised Scottish dentists to consider AP for high-risk individuals. We compared the effects of this differing guidance on AP prescribing and IE-admissions. Methods A quasi-experimental (natural) longitudinal-study design was used to compare aggregated national AP prescribing (2002–2025) and IE-admissions data (1998–2025) for England and Scotland. Interrupted time series and difference-in-differences analyses were used to evaluate changes. Findings AP prescribing fell 93% in England after the 2008 NICE guidelines and remained low (2567 [95% CI 2433–2701] prescriptions/million/annum before 2008 vs 185 [95% CI 165–204] prescriptions/million/annum after 2021/2, p < 0.0001), while IE-admissions increased by 112% (26 [95% CI 25–27] admissions/million/annum before 2008 vs 55 [95% CI 54–56]/million/annum after 2021/2, p < 0.0001). In Scotland, AP prescribing fell less, because GP prescribing replaced some dental prescribing, then rose again after 2018 and finished 2.86 times higher than in England (1884 [95% CI 1816–1952] prescriptions/million before 2008 vs 528 [95% CI 350–705] prescriptions/million after 2021/2). Correspondingly, Scottish IE-admissions rose less (24 [95% CI 23–25] admissions/million before 2008 vs 34 [95% CI 32–37] admissions/million after 2021/2, p < 0.0001) than in England, and English IE-admissions/million ended 61% higher than in Scotland (p < 0.0001). Interpretation Although our data do not establish causality, they suggest that cessation of AP may be followed by a significant increase in IE-admissions compared with policies recommending AP for high-risk individuals. Funding None.
REM sleep behavior disorder (RBD) is a well-established prodromal marker of α-synucleinopathies, including Parkinson’s disease, dementia with Lewy bodies, and multiple system atrophy. Over 80
